Reliability of Measurements Obtained
With a Modified Functional Reach
Test in Subjects With Spinal
Background and Purpose. The primary purpose of this study was to
determine whether the Functional Reach Test (FRT) could be modi-
fied to provide reliable measurements of sitting balance. A secondary
purpose was to determine whether the test could be used to measure
differences among levels of spinal cord injury. Subiects. Thirty male
subjects with spinal cord injuries were divided into three groups based
on injury type. Group 1 consisted of subjects with C5-6 tetraplegia,
group 2 consisted of subjects with T1-4 paraplegia, and group 3
consisted of subjects with T10-12 paraplegia. Methods. Subjects sat on
similar mat tables (tables varied based on what was available at a given
clinic) against the same backboard, set at 80 degrees. During two
sessions, forward reach was measured with a yardstick, with a 10-minute
break between sessions. Results. Intraclass correlation coefficients (3,2)
were high and varied from .85 to .94. Post hoc testing revealed that
differences occurred between groups 1 and 3 and groups 2 and 3, but
not between groups 1 and 2. Conclusion and Discussion. Test-retest
reliability was high with modification of the FRT with a single rater.
The measurements reflected differences among levels of lesion. Fur-
ther study is needed to determine normal values for all levels of lesion,
relationships to functional outcomes, and effects of equipment on
sitting balance. The modified FRT appears to provide reliable mea-
surements of sitting balance in nonstanding persons with spinal cord
injuries. [Lynch SM, Leahy P, Barker SP. Reliability of measurements
obtained with a modified Functional Reach Test in subjects with spinal
cord injury. Phys Ther. 1998;78:128-133.1
Key Words: Measurement, Reliability, Sitting balance, Spinal cord injuly.
Suzanne M Lynch
Susan P Barker
Ms LeaIiy died October 9. 1995
Physical Therapy . Volume 78 . Number 2 . February 1998
alance has been studied in various ways: by impairments, including persons with spinal cord injury
B recording biomechanical descriptions of bal-
ance reactions,'-4 by examining physiological
components of b a l a n ~ e , ~ - ~ by investigating
changes in the ability of a person to balance across the
life ~ p a n . ~ J OThese studies provide a basis for under-
The Functional Reach Test (FRT)Z0 can be used to
measure standing balance. In our view, the FRT is fast
and easy to use. A study using the FRT with 217 elderly
standing human performance. Often, the measurements male veterans (aged 70-104 years) demonstrated that
obtained. in research settings are not practical for rou- the test provides highly reliable measurements of bal-
tine clinical application. ance and can be used to predict the risk of falling.g1The
FRT also can be used to estimate physical frailty22 and to
Most studies of balance have been performed with demonstrate change in response to treatment.Z3 In the
subjects in the standing position, but studies of sitting study by Weiner et al,Z3 28 inpatient male veterans were
balance have also been reported.14-l7 Most studies of tested every 4 weeks during a regular physical therapy
sitting balance have used instrumentation similar to that program, and increases in functional reach and other
used for studies of standing balance.14J5 Some balance mobility measures were documented. No control was
tests that are less dependent on instrumentation have placed on the therapy received. Studies of FRT have also
been introduced, but these measures are designed for demonstrated strong reliability and ~alidity.2~-z3 The
persons who can ambulate.18J9Only a few tests exist for FRT, therefore, possesses attributes that can make it a
clinical balance assessment of nonstanding individuals. meaningful and accessible test.
One such test is the Seated Posture Control Measure,l"17
which is designed to document a child's posture in his or Measures that can be used to predict outcomes regard-
her seating system and to assess his or her ability to ing the balance of patients with SCI are not available.
function. Unfortunately, the test is quite long (36 items) Therapists cannot be certain that prescribed wheelchairs
and may not be generalizable to persons with a variety of or cushions provide patients with the most stable posi-
SM Lynch, PT, NCS, is Staff Physical Therapist, Brain Injury Rehabilitation Unit, St Mary's Hospital, Rochester, NY. She was Clinical Specialist,
Spinal Cord Injury, Magee Rehabilitation Hospital, Philadelphia, Pa, at the time of this study. This study was completed in partial fulfillment of
Ms Lynch's Master of Science degree in neurologic physical therapy at Philadelphia College of Pharmacy and Science, Philadelphia, Pa. Address
all correspondence to Ms Lynch at 39 Havenshire Rd, Rochester, NY I4625 (USA) (firstname.lastname@example.org).
P I.eal,y, PT, NCS, was Assistant Professor, Philadelphia College of Pharmacy and Science, at the time of this study.
SP Barker, PT, NCS, is Associate Professor, Philadelphia College of Pharmacy and Science.
This study was approved by the Human Subjects Review Board of Magee Rehabilitation Hospital and Philadelphia College of Pharmacy and
This work is adapted from a platform presentation at the Combined Sections Meeting of the American Physical Therapy Association; February
8-12, 1995; Reno, Nev.
This adick was submitted Sqbtember 9, 1996, and was accepted June 20, 1997.
Physical Therapy . Volume 7 8 . Number 2 . February 1998 Lynch et a l . 129
tions from which they can function (ie, the best bal- and the presence of musculoskeletal deformities in the
ance). Defining positions that are stable and the effects upper extremity used in reaching were examined at the
of equipment on stability would be helpful because time of the testing. The presence of inadequate muscle
persons with paralysis are challenged to maintain their force to maintain shoulder flexion during reaching (as
balance during a variety of functional activities. measured by a break test of the shoulder flexors),
inadequate range of motion, or musculoskeletal defor-
For the purposes of our study, we defined sitting balance mity meant elimination from the study. Spasticity, a
as the ability of a person to maintain control over common sequela in persons with SCI, was not part of the
upright posture during forward reach without stabiliza- inclusion or exclusion criteria. Spasticity was not mea-
tion. Any reaching task will be a challenge to upright sured in any subjects.
control for persons with partial or complete paralysis of
the trunk and arms. The primary purpose of our study Instrumentation
was to determine whether the FRT could be modified for A yardstick was attached horizontally to a wall by Vel-
a group of individuals with SCI to provide reliable cro@* or tape. The method of attachment varied,
measurements of sitting balance. A secondary purpose depending on the site of data collection. According to
was to determine whether the modified FRT could Duncan et the method used to attach the yardstick
measure differences in functional reach among different is not crucial. All subjects sat on a narrow mat table or a
levels of SCI. padded weight bench, which were of similar width
(about 61 cm [24 in]). The same backboard was used and
Method kept at the same angle of 80 degrees for all subjects. This
angle allowed all subjects to sit back and relax between
Subjects trials. The backboard used in this study is also typically
Thirty male subjects participated in this study. The used for supporting sitting activities during rehabilita-
subjects were between 18 and 45 years of age (X=30.8, tion of patients with SCI (Figure).
SD=7.2). Subjects were placed in groups based solely on
level of injury. All subjects had complete lesions accord- Procedure
ing to the American Spinal Injury Association's (ASIA) Informed consent was obtained once subjects were
Impairment S ~ a l e . 2 ~ lesions, therefore, were classi-
The determined to be eligible for the study. The procedure
fied as either ASIA A or ASIA B, because both classifica- for the collection of data closely followed the procedure
tions are for complete motor injuries. The difference described by Duncan et al.2O Once each subject was
between the categories is in sensation. There is no positioned on the mat table, the yardstick was placed
sensation below the level of the lesion in ASIA A lesions, along the subject's shoulder at the level of the acromion.
but sensation can be partially spared in ASIA B lesions. Subjects sat in the same position for each trial. Their
We chose these type categories of lesions to ensure that hips, knees, and ankles were positioned with 90 degrees
there would be no lower-extremity motor function to of flexion, and there was 5.08 cm (2 in) of clearance
allow the subject to weight bear on the feet when between the popliteal fossa and the mat table. Foot
reaching forward in sitting. All subjects were recruited support was provided, if necessary, with a rubber floor
from the following sources: scheduled medical appoint- mat to ensure proper sitting position. The backboard
ments for Magee Rehabilitation Hospital's (Philadel- was placed behind each subject for support (Figure).
phia, Pa) SCI follow-up system, teams that participated
in wheelchair sports tournaments, and persons readmit- Initial reach was measured with each subject resting
ted to Magee Rehabilitation Hospital for intensive reha- against the backboard with an upper-extremity flexed to
bilitation. All subjects were seen at least 1 month after 90 degrees. The anatomical landmark used to measure
completion of their initial phase of rehabilitation. Sub- reach was the ulnar styloid process. Because the subjects
jects were selected based on their SCI diagnosis and with tetraplegia in our study could not make a fist, this
assigned to one of three groups: group 1 (n=10) con- landmark was used instead of the third metacarpal,
sisted of subjects with C5-6 tetraplegia, group 2 (n=10) which was used in the original studies of FRT.20-23 The
consisted of subjects with T1-4 paraplegia, and group 3 ulnar styloid process is a prominent landmark and was
(n=10) consisted of subjects with T10-12 paraplegia. proximal enough to allow accurate measurements to be
taken for all subjects. Subjects used the nonreaching
To be included in our study, subjects had to be able to sit upper extremity for counterbalance only (eg, no weight
independently of a seating system with only a backboard bearing or holding on was allowed). The subjects were
for support. The subjects' upper extremities had to be guarded for safety, and the trial was repeated if the
without deformities, and each subject had to be able to subject required assistance to recover to the backboard.
assume and maintain 90 degrees of shoulder flexion.
Muscle force (manual muscle testing), range of motion,
'Velcro USA Inc, 406 Brown Ave, Manchester, NH 03108.
130 . Lynch et al Physical Therapy. Volume 7 8 . Number 2 . February 1998
Two sites were used for data collection. Limitations of
the physical facilities at one of the data collection
locations necessitated that all 8 subjects who were tested
there use their left upper extremity. The remaining 22
subjects who were tested at the other facility used their
right upper extremity. All methods were otherwise the
same between the sites.
a .I. r
Each subject had two practice trials of maximal forward
reach, followed by three trials during which data were 4 9 '
collected. The mean of these three trials was recorded. 3 <
Following the initial three trials, each subject left the
testing area for 10 minutes and then returned to
undergo repeated testing using the same procedure. A
single rater (SML) collected all data for this study.
Test-retest reliability was studied using the intraclass
correlation coefficient (ICC[3,2]) because there was a
single rater." Calculations were performed using a
spreadsheet software package.t Because a secondary
purpose of our study was to determine whether the
modified FRT could measure differences among levels
of lesion, a one-way analysis of variance (ANOVA) was
used to test for differences among the means for reach
in the three groups. A Newman-Keuls test was used to
discern differences among group means and to ensure
that Type I error was minirni~ed.?~ Figure.
Subject demonstrating positioning with backboard for data collection.
Mean reach data for the subjects are presented in the
Table. The results indicated that the reliability of mea- Discussion
suremen1.s obtained with the modified FRT was very Forward reach in a sitting position can be measured
strong. The ICCs for test-retest reliability of measure- reliably via a ruler attached to a wall alongside a patient
ments of average reach length were .94 for group 1, .85 with SCI. The modified FRT achieved ICCs for test-retest
for group 2, and .93 for group 3. There were no reliability similar to those documented in the original
differences between the subjects who used their right ~-~"
FRT s t u d i e ~ . ~ Generalizability of test-retest reliability
upper extremity and the subjects who used their left is weak because only one rater was available for data
upper extremity to perform the reaches. collection. Further study using an interrater design may
allow inferences to be generalized to a greater number
The modified FRT was also tested for its ability to of situations.
distinguish level of lesion. Mean maximal reach was 14.7
cm (SD-7.6, range=3.3-27.4) for group 1, 15.5 cm The modified FRT appears to be useful for determining
(SD=4.3, range=7.6-21.3) for group 2, and 22.9 cm differences in reach among different levels of lesion in
(SD=5.6, range= 14.7-29.2) for group 3. A one-way persons with SCI. The modified FRT measured differ-
ANOVA was used to determine that subjects with lower ences in reach between groups 1 and 3 and groups 2 and
levels of lesion had a longer reach compared with 3. There was no difference in the ability to reach
subjects with higher levels of lesion. The Neuman-Keuls between groups 1 and 2, but mean reach was greater in
test demonstrated that reach differed only between group 3 compared with groups 1 and 2. This finding
groups 1 and 3 and groups 2 and 3. There was no appears to be reasonable because people with lower
difference in reach between groups 1 and 2. levels of paraplegia tend to have greater functional
capabilities than people with higher levels of lesion do.
The subjects in group 3 had abdominal and back exten-
sor muscles that were unaffected by their SCI, which
apparently gave them a greater advantage in movement
t Microsoft Excel 5.0, hficrosoft C:orp, One Microsort Way, Redmond, WA 98052.
Physical Therapy Volume 78 . Number 2 . February 1998 Lynchetal. 131
Table patients are challenged appropriately by a test, and
Maximal Functional Reach
poorer performances occur when patients believe that
the test has no meaning for their problem. Face validity
Maximal appears to be present in the modified FRT because
No." Reach ( ~ m ) ~ subjects felt the challenge to their stability and had to
make great effort not to fail or a fall would occur.
1 14.10 (5.55)
2 8.48 (3.34) Future research is needed to obtain evidence that the
3 14.61 (5.75)
4 27.53 (1 0.84) modified FRT can be used to predict future outcomes
5 19.91 (7.84) (predictive validity) or current balance status. We believe
6 22.12 (8.71) that the modified FRT should be compared with mea-
7 3.40 (1.34) sures of established criterion-related validity. Strength-
8 12.70 (5.00) ening validity may demonstrate that the modified FRT is
9 2.54 (3.21)
10 15.04 (5.92)
a proper method to answer clinical or research
11 17.78 (7.00) questions.
12 15.77 (6.21)
13 7.62 (3.00) Studies using the modified FRT would improve its
14 20.32 (8.00) usefulness. Because patients with SCI sit on different
15 2 1.39 (8.42)
16 1 1.63 (4.58) support surfaces (cushions and wheelchairs), compari-
17 12.93 (5.09) sons could be made only among different products.
18 17.35 (6.83) Measurement of functional reach may cause clinicians to
19 12.80 (5.04) prescribe equipment based on its effects on sitting
20 17.48 (6.88) balance.
21 29.12 (1 1.46)
22 27.3 1 (10.75)
23 19.91 (7.84) Conclusion
24 24.03 (9.46) This study examined the use of the FRT for a population
25 17.15 (6.75) that cannot stand: persons with complete SCI. The
26 28.37 (1 1.17) modified FRT can become a highly useful test because it
27 28.37 (1 1.17)
28 23.93 (9.42) is easy and fast to perform and adaptable to many
29 15.24 (6.00) environments. The purpose of this study was to test
30 14.71 (5.79) whether the FRT could provide reliable measurements
"Group 1 (C.5-6 tetreplegia): subjects 1-10; group 2 ( T l 4 paraplegia):
in persons who are unable to stand. Before the measure-
subjects 11-20; group 3 (TlG12 paraplegia): subjects 21-30. ments can be shown to be useful, research on their
"Measurements in inches shown in parentheses. validity is needed.
The modified FRT did not appear to detect differences 1 Kirby RL, Price NA, MacLeod DA. The influence of foot position on
between the subjects with tetraplegia (group 1) and the standing balance. J Biomech. 1987;20:423-427.
subjects with higher levels of paraplegia (group 2). 2 Mahar RK, Kirby RL, MacLeod DA. Simulated leg-length discrep
Although the subjects with higher levels of paraplegia ancy: its effect on mean center of pressure position and postural sway.
had more unaffected muscles than the subjects with Arch Phys Med Rehabil. 1985;6:822-824.
tetraplegia did, reach outcomes were similar. Further 3 Nasher LM, Cordo PI. Relation of automatic postural responses and
study is needed. reaction-time voluntary movement of human leg muscles. Exp Brain
Although our study indicates that reliability exists for 4 Alexander NB, Shepard N. Gu hlJ, Schultz A. Postural control in
measurements obtained with the modified FRT, more young and elderly subjects when stance is perturbed: kinematics.
research is needed to establish validity. Face validity is J Gerontol. 1992;47:M79-M87,
the assessment of how well a test appears to measure 5 Diener HC, Horak FB, Nasher LM. Influences of stimulus parameters
something specific. In our study, subjects with varying on human postural responses. J Neurqhysiol. 1988;59:1888-1905.
amounts of paralysis were asked to reach forward and 6 Dichgans J, Diener HC. The contribution of vestibulo-spinal mech-
move without any assistance from their base of support. anisms to the maintenance of human upright posture. Actn Otolalyngol
We believed that each subject had to move to the limits
of his stability without loss of balance. We contend that 7 Magnusson M, Johansson R. Dynamic performance of vibration
it is important for a test to measure what clinicians and induced anterior-posterior sway during upright in normal subjects.
Acla Ololalyngol (Stockh). 1990;llO:168-1 74.
patients believe can affect the patients' functional per-
formance. According to Campbellz6in her discussion of 8 Hasselkus BR, Shambus GM. Aging and postural sway in women.
J Gerontol. 1975;30:661-667.
face validity, better performances may occur when
132 . Lynch et a1 Physical Therapy . Volume 78 . Number 2 . February 1998
9 Black FO, Wall C, Rockette HE, Kitch R. Normal subject postural 18 Tinetti M. Performance-oriented assessment of mobility problems
sway during the Romberg test. Am J Ololalyngol. 1982;3:309-318. in elderly patients. J A m Gen'alr Soc. 1986;34:119-126.
10 Fernie GR, Gryfe CI, Holliday PJ, Llewellyn A. The relationship of 19 Shumway-Cook A, Horak FB. Assessing the influence of sensory
postural sway in standing to incidence of falls in geriatric subjects. Age interaction o n balance: suggestion from the field. Phys Ther. 1986;66:
Ageing. 1982;ll:ll-16. 1548-1550.
11 Horak FB, Shupert CL, Mirka A. Components of postural dyscon- 20 Duncan PW, Weiner DK, ChandlerJ, Studenski S. Functional reach:
trol in the elderly: a review. Neurobiol Aging. 1989;10:727-738. a new clinical measure of balance. J Gerontol. 1990;45:M192-M197.
12 Woollacott M. Postural control mechanisms in the young and old. 21 Duncan PW, Studenski S, Chandler J, Prescott B. Functional reach:
In: Duncan PW, ed. Proceedings of the APTA Fonrm on Balance; June predictive validity in a sample of elderly male veterans. J Gerontol.
13-15, 1989; Nashville, Tenn. Alexandria, Va: American Physical Ther- 1992;47:M93-M98.
apy Association; 199023-28.
22 Weiner DK, Duncan PW, ChandlerJ, Studenski S. Functional reach:
13 Kleiber M, Horstmann GA, Dieh V. Body sway stabilization in a marker of physical frailty.J A m Genatr Soc. 1992;40:203-207.
human posture. Acta Otolulyngol (Stockh). 1990;110:168-174.
23 Weiner DK, Bongiorni DR, Studenski S, et al. Does functional reach
14 McClenaghan BA. Sitting stability of selected subjects with cerebral improve with rehabilitation? Arch Phys Med Rehabil. 1993;74:796-800.
palsy. Clin Biomech. 1989;4:213-216.
24 Maynard FM. Intnnational Standards for Neurological and Functional
15 Reid D T , Sochaniwskyj A, Milner M. An investigation of postural Classification of .Spinal Cord Injuly. Chicago, Ill: American Spinal Injury
sway in sitting of normal children and children with neurological Association; 1996:7.
disorders. Physical 6 Occupational Therapy in Pediatrics. 1991;11:19-34.
25 Portney LG, Watkins MP. Validity of measurements. In: Foundations
16 Fife SE, Roxborough LA, Armstrong RW, et al. Development of a of Clinical Research: Applirations lo Practice. East Norwalk, Conn: Apple-
clinical measure of postural control for assessment of adaptive seating ton & Lange; 1993:69-85.
in children with neuromotor disabilities. Phys Ther. 1991;71:981-993.
26 Campbell SK. Commentary on "Measurement validity in physical
17 Fife SE, Roxborough LA, Story M, et al. Reliability of the Seated therapy research." Phys Ther. 1993;73:60-61.
Posture Control Measure. Presented at the Ninth International Seating
Symposium; February 1993; Memphis, Tenn.
Physical Therapy . Volume 78 . Number 2 . February 1998 Lynch et al . 133